Re: [boost] functional.hpp: Problem with not1, not2

From: Mark Rodgers (mark.rodgers_at_[hidden])
Date: 2000-11-20 02:48:02

From: "Jens Maurer" <Jens.Maurer_at_[hidden]>

> In short, I believe Comeau is correct and all declarations of not1, not2
> in functional.hpp should be changed to
>
>   template<class Pred>
>   void functor(Pred p);
>
> instead.
>
> What's your opinion?
>
> Jens Maurer

I believe you are right that Comeau is correct.  However, I would prefer
not to pass by value in case the functor is of not trivial size, so wonder
if an overload would be appropriate:

    template <class Predicate>
    unary_negate<Predicate> not1(const Predicate &pred)
    {
        // ... as now
    }

    // new
    template <class Predicate>
    unary_negate<Predicate> not1(Predicate &pred)
    {
        return unary_negate<Predicate>(pred);
    }

What does Comeau make of this?  BCC5.5 and GCC 2.95.2 seem to be happy
that it doesn't introduce any ambiguity.

Mark